1. Field of the Invention
The present invention relates to an image scaling method and apparatus, and more particularly, to an image scaling method and apparatus in a digital television.
2. Description of the Prior Art
As conventional analog video broadcasting is transforming to digital video broadcasting (DVB), digital televisions become more popular.
The digital television signals include video data and caption data, wherein the caption data is generally a subtitle or a closed caption such as scripts, logos, scrolling banners, etc. Typically, different circuits process the video data and the caption data respectively. The processed video data, processed caption data, and other information are then mixed by a video mixer/blender to form a composite frame.
When a user enables a television function that changes the size of the pictures (e.g., the picture-in-picture (PIP) function), the sizes of both the video data and the caption data must be scaled. Well known in the art, the foreground pixels and the background pixels of the caption data are represented with a pixel format containing a transparency parameter (e.g., an α value) such as the ARGB or AYCbCr format during the scaling operations. The foreground pixel and the background pixel have distinct α values, so, the following stages such as the video mixer/blender, can differentiate the foreground and the background of the caption accordingly.
The conventional art typically scales the caption data by interpolation. Unfortunately, gradient colors, sawtooth phenomenon, and other undesirable problems usually appear on the edge of the foreground of the scaled caption and thereby deteriorate the smoothness and sharpness of the scaled caption data.